Let's talk

AI Architect & Founder

Divyaprakash Jha

An agentic AI architect based in Palghar, Maharashtra, building the future of decision intelligence and business automation.

By Divyaprakash JhaUpdated 5 min readTopic: Founder

Photo for Divyaprakash Jha

Who is Divyaprakash Jha?

Divyaprakash Jha is an agentic AI architect and the founder of GodEngine and AskShiva. Based in Palghar, Maharashtra, he builds AI-native operating systems and decision intelligence engines designed to automate complex business workflows.

He operates a specialized studio that functions with a unique human-to-AI ratio. The team consists of three human operators who direct a fleet of fifteen AI agents to develop and ship software products.

  • Founder: Leads GodEngine and AskShiva.
  • Specialization: Agentic AI architecture and decision intelligence.
  • Location: Palghar, Maharashtra, India.
  • Operational Model: A studio of three humans and fifteen AI agents.

From Trading Systems to AI Architecture

Divyaprakash's journey into complex system design began in the financial markets. He spent ten years working in the Gold and FX markets, starting his trading journey in 2016. He learned firsthand that markets are unforgiving and require disciplined, rule-based execution.

This focus on discipline led him to move from manual trading to building automated systems. He developed over 500 Expert Advisors (automated trading programs) for MetaTrader 5. Through his previous ventures, including CandleForge and WICKFURY, he served approximately 100 clients, providing custom builds and subscription-based software for trading desks.

In April 2026, he transitioned to full-time, independent AI development, applying the logic of "small agents" (Expert Advisors) to the world of Large Language Models. He views the jump from trading automation to AI architecture as a natural evolution of the same core principle: building systems that read signals, decide, and act without constant human intervention.

GodEngine: Decision Intelligence

Most AI tools are designed to answer questions, but GodEngine is designed to help you make decisions. It is a decision intelligence engine that moves beyond the simple "chat" interface to provide structured, auditable reasoning.

When you pose a decision question—such as whether to open a new shop in Vasai or invest in delivery—GodEngine does not just give a fluent response. It treats the question as a model of the world. It extracts facts, identifies cause-and-effect links, and builds a knowledge graph that you can interact with.

Key Features of GodEngine

  • Scenario Probing: It builds multiple plausible futures and tests what would have to be true for each scenario to occur.
  • Reasoning and Citations: Every claim in a report points back to a specific entity, link, or source, allowing you to audit the logic.
  • Streaming Pipeline: The engine uses server-sent events so you can watch the reasoning process happen in real-time rather than waiting for a static block of text.

The engine is available on the Web and via an Android app, providing the same interactive graph experience across both platforms.

AskShiva: AI for Indian Business

While GodEngine focuses on high-level strategy, AskShiva is built for the practical, daily tasks of running a business. It is an AI assistant specifically designed for the Indian context, recognizing that local business relies on specific workflows like WhatsApp, GST, and UPI.

AskShiva bridges the gap between "having an answer" and "getting the work done." It is built to handle multi-step tasks that require memory across different conversations.

Practical Business Examples

Instead of just explaining how to do a task, AskShiva can execute it. For example:

  • Invoicing: "Make a GST bill for Sharma Traders, ₹18,400." The assistant drafts the file with the correct tax lines.
  • Communication: "Send it on WhatsApp, in Marathi." It translates the message and prepares it for sending.
  • Follow-ups: Because it remembers the customer and the transaction, it can handle reminders like, "Remind me if they have not paid by Friday."

The platform supports voice interaction in the browser and is designed to work with the reality of Indian business, including local language translations and standardized document formats.

The Agentic Studio Model

The Development Loop

The studio follows a strict "plan, act, verify" loop for every product. The process involves:

  1. Intent and Constraints: Humans define the high-level goal and the boundaries.
  2. Parallel Building: AI agents build components in parallel to increase speed.
  3. Early Attack: Every idea is "attacked" early in the process to find flaws.
  4. Verification: The system verifies the output against the original intent.

To maintain independence and control costs, the studio runs its own proprietary GPU and CPU servers using open models. This ensures that bulk AI work does not depend on expensive, paid third-party APIs.

Shiva Agents & Workflow Automation

For organizations looking to scale AI beyond simple chat, Shiva Agents provides a platform to run AI agents as an "always-on" company. This is not just a demo loop; it is a production-grade system designed for reliability and governance.

Users can build agents in a "Agent Studio" by defining their role, goal, and style, or by simply describing what they want in plain language. These agents can work alone or in complex teams to complete workflows.

Feature Description
Visual Workflows A canvas-style interface (similar to n8n) where nodes represent agents, teams, or logic branches.
Autonomy Levels Control settings ranging from "Observe" to "Assist" and "Act" to "Autonomous."
Governance Includes audit logs, approval gates for risky steps, and time-travel capabilities to fork runs.
Durability Supports scheduled "heartbeats" (cron jobs) and can resume work after a server restart.

The platform supports 23 runnable workflow patterns and 20 different node types, allowing for highly sophisticated automation.

BusinessOS: The AI-Native OS

Most business software is a collection of disconnected tools—one for CRM, one for HR, and another for finance. This creates "seams" where data is duplicated or lost. BusinessOS is an AI-native operating system designed to eliminate these silos by providing a single, unified foundation.

It is a multi-tenant platform that centralizes essential business functions under one identity and permission engine. This ensures that data is isolated at the database level and every action is recorded in an append-only audit log.

The Foundation First Approach

Instead of bolting modules together, BusinessOS starts with a master architecture. Currently, the live foundation includes:

  • Secure Access: Support for passkeys, two-factor authentication, and SSO (Google/Microsoft).
  • Organization Management: Tools to manage different departments, teams, and workspaces.
  • Data Integrity: Row-level security (RLS) and a single source of truth for all business data.
  • Search & Notifications: A unified system for finding files and receiving updates across the organization.

By providing a single permission engine and a shared event bus, BusinessOS allows businesses to eventually add modules like CRM or Finance on top of a rock-solid, secure core.

Questions people ask

What companies did Divyaprakash Jha found?

He founded GodEngine, a decision intelligence engine, and AskShiva, an AI assistant for business automation.

Where is Divyaprakash Jha based?

He is based in Palghar, Maharashtra, India.

What is the difference between GodEngine and AskShiva?

GodEngine is a decision intelligence engine that probes scenarios and shows reasoning, while AskShiva is an assistant designed to complete practical business tasks and workflows.

How many people work at his studio?

The studio consists of three human operators directing a fleet of fifteen AI agents.

What is the focus of his AI work?

He focuses on agentic architecture, building systems that can plan, act, and verify tasks autonomously.